Correct me if I'm wrong, but this methodology seems flawed. Using building permits per 1000 residents only provides an accurate comparison metric if the density

Correct me if I'm wrong, but this methodology seems flawed. Using building permits per 1000 residents only provides an accurate comparison metric if the density of the housing is the same. If the move is (as it seems to be) towards large multi-unit housing complexes, shouldn't we be looking at residential units created?